Traffic contribution to PM_2.5 increment in the near-road environment

Highlights

• We highlight the importance of methodology used to estimate background concentration in transportation air quality studies.

• The traffic contribution to 24-hour PM_2.5 increment in the studied near-road environment is 23% of background concentration.

• The effect of wind speed and wind direction on near-road traffic-related air pollution is significant.

Abstract

A growing number of studies have reported the adverse health effects of long-term exposure to air pollutants, especially fine particulate matter (PM_2.5). Vehicular emission sources have been shown to contribute to elevated air pollution concentrations in the near-road environment, including PM_2.5, based on monitoring data collected mainly during short-term campaigns. The United States Environmental Protection Agency (EPA) added near-road monitors to its national network to collect long-term National Ambient Air Quality Standard (NAAQS)–comparable data in the near-road environment. The EPA also mandated inclusion of near-road monitoring data in the Air Quality Index to reflect the elevated level of near-road PM_2.5 concentrations to which millions of people in major urban areas are exposed to on a daily basis. For the first time, PM_2.5 data collected at one of these near-road monitors were compared with those of other NAAQS monitors during 2016 in Houston, Texas. One of these NAAQS monitors was selected based on EPA guidance for quantitative hotspot analyses of particulate matter to represent background concentrations. The near-road PM_2.5 increment was statistically significant. The traffic contribution to 24-h PM_2.5 increment in the near-road environment was estimated to be about 23% of background concentration, which is close to estimates given by previous studies (22%) and is greater than a recent estimate based on a national-scale data analysis (17%), emphasizing the importance of background monitor selection criteria. Wind speed and direction were shown to have a considerable effect on PM_2.5 increment in the near-road environment. A multiple linear regression model was developed to predict 24-h near-road PM_2.5 concentrations using background PM_2.5 concentration, wind speed, and wind direction. This model explained 83% of the variability of 24-h PM_2.5 concentrations in the near-road environment and showed improvement in near-road concentration predictions when accounting for wind speed and wind direction.

Introduction

Adverse health effects of exposure to fine particulate matter (PM_2.5) have been investigated in a growing number of studies (Bell et al., 2010; Khreis et al., 2017; Sapkota et al., 2012, Sohrabi et al., 2019; Sharifi et al., 2019). Exposure to an increased level of PM_2.5 concentrations has been associated with adverse health effects such as increased blood pressure and hypertension, increased rates of ischemic stroke, and narrower arterial diameter (Foraster et al., 2014; Qiao et al., 2014; Wellenius et al., 2012; Zamora et al., 2018). Recent studies have also revealed strong evidence of the relationship between long-term exposure to PM_2.5 and common neurodegenerative diseases (Kioumourtzoglou et al., 2016). Further studies also have shown a significant association between an increase in traffic-related PM_2.5 and diseases like cardiac anomalies (Girguis et al., 2016). The increment of traffic-related PM_2.5 in the near-road environment can potentially yield a series of adverse health effects for a large number of people who reside near roadways (Weinstock et al., 2013). Rowangould (2013) reported that more than 19% of the United States (US) population lives within 100 m of a high-volume roadway. According to the 2013 national household survey, 16.88 million households lived within half a block of a major transportation facility in 2011, yielding exposure of more than 40 million people to an elevated level of PM_2.5 (U.S. Department of Housing and Urban Development, 2016). Many people throughout the world also live near major roadways, so the worldwide population exposed to an elevated level of traffic-related air pollutants including PM_2.5 is potentially much larger.

The US Environmental Protection Agency (EPA) added a number of near-road monitors to its network and mandated inclusion of near-road monitoring data in the Air Quality Index (AQI) to reflect the potential for an elevated level of near-road PM_2.5 concentration to which millions of people may be exposed on a daily basis in a considerable portion of major urban areas (U.S. EPA, 2013). One of the key objectives of this program was to collect National Ambient Air Quality Standards (NAAQS)–comparable datasets from the near-road environment as input to studies on adverse health effects of long-term exposure to PM_2.5. The appropriate monitoring methods for this purpose are the Federal Reference Method (FRM), Federal Equivalent Method (FEM), and Approved Regional Method (ARM), despite some limitations of these methods (U.S. EPA, 2009, 2013). EPA guidance (EPA, 2012) requires a probe for near-road monitoring to be located as close as possible to and not farther than 50 m from the outside nearest edge of the road and between 2 and 7 m in height from the road elevation. It also requires PM_2.5 state and local air monitoring stations “to operate on at least a 1-day-in-3 sampling schedule” to be able to characterize elevated PM_2.5 concentration near heavily traveled roads (U.S. EPA, 2013).

Characterization of near-road traffic-related air pollutants has been performed in a variety of studies including tracer studies (Benson, 1989; Cadle et al., 1977; Finn et al., 2010; Venkatram et al., 2013), short-term field studies (Baldauf et al., 2008; Klompmaker et al., 2015; Patton et al., 2017; Venkatram et al., 2013; Zhu et al., 2009), intensive field studies (Ginzburg et al., 2015; Kimbrough et al., 2013; Zamora et al., 2018; Zhang et al., 2017), and modeling (Askariyeh et al., 2017, 2018, 2019; Vallamsundar et al., 2016; Heist et al., 2013; Isakov et al., 2014; Steffens et al., 2014). These studies have reported that traffic-related air pollution depends on wind speed and direction; peaks at the nearest points to the roadway; decreases exponentially with distance from the road; and, reaches the background concentration over a distance of a few hundred meters. A synthesis of previously collected real-world data showed a 22% increment of PM_2.5 in the near-road area compared with background PM_2.5 concentration (Karner et al., 2010), while other studies have estimated this value to be 13%–20% (Ginzburg et al., 2015), and 10%–17% (DeWinter et al., 2018; Guerreiro et al., 2011; Keuken et al., 2013). It should be noted that a recent national-scale review of near-road concentrations provided the average near-road PM_2.5 increment in the US and did not investigate the Houston monitor because the methodology completeness criteria were not met for background monitor selection (DeWinter et al., 2018).

A key factor in understanding the traffic contribution to the near-road increment of air pollution is the background concentration estimate when speciation monitor analysis (Ginzburg et al., 2015; Sofowote et al., 2018) is not available. EPA guidance (2015b) for quantitative hotspot analyses of particulate matter describes how to determine background concentrations from sources other than the study target. It describes how to use the data of a single monitor with similar characteristics, that is close in proximity, and that is located upwind of a target area to be as representative as possible for the background concentration. This guidance also explains how to use a wind rose to identify upwind sources and how to use inverse distance weighting to interpolate among several monitors if no single monitor represents the background concentration (U.S. EPA, 2015b). Researchers have used concentrations observed at a distance of a few hundred meters from the target roadway and the first percentile of hourly near-road monitoring (Patton et al., 2017), as well as concentrations monitored at a location not affected by any close emission sources and a high correlation with target monitor (DeWinter et al., 2018), to determine background concentrations and estimate traffic-related PM_2.5 increment in the near-road environment. Considering different methods and datasets, mainly obtained from short-term campaigns due to the rare availability of long-term near-road monitoring, range of values have been reported to quantify the traffic contribution to PM_2.5 increment in the near-road environment.

The main objective of this study was to quantify PM_2.5 increment due to traffic contributions in a near-road environment at different wind speeds and wind directions using NAAQS-comparable near-road monitoring data. To accomplish this objective, the 24-h (daily average) PM_2.5 concentrations monitored at a near-road location, collected as part of the EPA near-road monitoring network (U.S. EPA, 2013), as well as data from all other NAAQS monitoring stations during 2016 in Houston, were compared and analyzed. The monitor representing background PM_2.5 concentrations was selected to estimate the PM_2.5 increment due to traffic in a near-road environment. The PM_2.5 increment due to traffic was evaluated under different wind speeds and wind directions, and a multiple linear regression model was developed to predict the near-road 24-h PM_2.5 concentration. To the knowledge of the authors, this is the first time that near-road PM_2.5 concentrations observed by a NAAQS monitor in Houston were investigated to quantify the traffic contribution to near-road PM_2.5 increment. The results of this study will give researchers a better understanding of the effect of transportation sectors on near-road PM_2.5 concentrations based on long-term observations that can be used for exposure assessments.